Abstract
Objective Psoriatic arthritis (PsA) is an immune-mediated disease characterized by pain, stiffness, and swelling of peripheral joints, with an estimated prevalence in Canada of 0.45%. Treatment aims to minimize disease activity, reduce progression of damage, and improve quality of life. Secukinumab (SEC) is a biologic disease-modifying antirheumatic drug (bDMARD) that has demonstrated efficacy and safety for PsA in clinical trials; however, there is limited real-world evidence on its use in Canada. The objective of this study was to use the Canadian Spondyloarthritis (CanSpA) Research Network to describe real-world retention and effectiveness of SEC among Canadian patients with PsA.
Methods This was an observational cohort study of Canadian patients with PsA, 18 to 65 years of age, who attended a clinic of the CanSpA network and received treatment with SEC. Patients were indexed on the date they first initiated SEC. Retention was assessed at 12 months postindex. Clinical effectiveness was measured as the proportion of patients in remission and change in disease activity from baseline to 12 months using several clinical indices.
Results In total, 213 patients were included. Overall retention was estimated at 73.6% at 12 months (81.8% for bDMARD- or targeted synthetic DMARD-naïve patients). Out of 110 patients, 17 (15.5%) were in remission based on the Disease Activity Index in Psoriatic Arthritis in 28 joints, and 10 out of 70 patients (14.3%) were in remission based on the Psoriatic Arthritis Disease Activity Score at 12 months. The Psoriasis Area and Severity Index improved by 65.8%; the tender joint count in 68 joints and the swollen joint count in 66 joints improved by 65.5% and 73.7%, respectively.
Conclusion This is the first nationwide study that we know of to describe real-world use of SEC in Canada for PsA, and the results support its effectiveness in a Canadian real-world setting. The CanSpA network represents a unique opportunity to build and improve the real-world evidence base for SpA treatment in Canada.
Spondyloarthritis (SpA) is described as a group of immune-mediated rheumatic diseases characterized by inflammation in the spine and joints.1 Psoriatic arthritis (PsA) is 1 disease subtype within the SpA family, characterized by 6 key disease domains: peripheral arthritis, axial disease, enthesitis, dactylitis, skin psoriasis (PsO), and nail PsO.1,2 Pattern of disease manifestations will vary from patient to patient, but typically present as pain, stiffness, and swelling of peripheral joints, predominantly, as well as skin PsO.1 PsA imparts a significant physical and psychological burden to patients that often leads to impaired function and disability and reduced quality of life.1 The prevalence of PsA among Canadians has been estimated to be up to 0.45% of the population,3 and PsA is more common among those with PsO, for whom the prevalence of PsA rises to almost 30%.4 Recommendations and clinical guidelines on the management of PsA have been published by the Canadian Rheumatology Association and the Spondyloarthritis Research Consortium of Canada in 2015,5 the European Alliance of Associations for Rheumatology in 2019,6 and the American College of Rheumatology in 20187; however, new therapeutic options for PsA have since emerged. The Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) presented a 2021 update to the recommendations for PsO and PsA to reflect this recent progress in treatment.6 Currently, first-line therapy options among treatment-naïve patients with PsA include conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), such as the small molecule methotrexate; biologic DMARDs (bDMARDs), such as tumor necrosis factor inhibitors; interleukin 17A (IL-17A), IL-23, and IL-12/23 inhibitors; and targeted synthetic DMARDs (tsDMARDs), such as Janus kinase inhibitors.5,7,8 In selecting initial treatment, GRAPPA recommendations emphasize that choice of therapy for PsA should consider all disease manifestations present in the patient, with the optimal therapy being the one that treats the majority of those manifestations.6 However, reimbursement criteria for most bDMARD and tsDMARD agents in Canada requires failure of csDMARDs before b/tsDMARDs can be initiated, thus limiting the physician’s choice of first-line therapy.9,10 If PsA is active despite first-line therapy, bDMARDs or tsDMARDs may be initiated.5
The bDMARD secukinumab (SEC) is an IL-17A inhibitor suppressing the inflammatory immune response; SEC received approval from Health Canada in April 2016 for the treatment of PsA in adults experiencing inadequate response to previous DMARD therapies.11 It has demonstrated efficacy and safety for use among patients with PsA through a clinical trial program accumulating up to 5 years of data, where patients showed significant improvements in all PsA core domains of the GRAPPA-Outcome Measures in Rheumatology (GRAPPA-OMERACT) set, including musculoskeletal and skin disease activity, pain, patient global assessment, physical function, health-related quality of life, fatigue, and systemic inflammation.12-15 Clinical effectiveness of SEC has also been demonstrated in a real-world setting outside of Canada. In Europe, real-world data collected through the European Spondyloarthritis (EuroSpA) Research Collaboration Network showed remission rates of up to 39% and a retention rate of 76% at 12 months postinitiation of SEC.16,17 However, there is a paucity of evidence on real-world use of SEC in Canada for the treatment of PsA.
The Canadian Spondyloarthritis (CanSpA) Research Network was created to leverage collaboration between several independent Canadian SpA registries. The aim of the CanSpA Research Network is to generate comprehensive real-world SpA evidence to answer questions for which a larger landscape coverage represents an advantage. Essentially, patient data collected during routine clinical practice across the participating registries is deidentified, pooled in a single location, and synthesized. The current study describes one of the first analyses conducted using data from CanSpA to address the aforementioned research gap. The aim of this study was to assess real-world retention of SEC and its clinical effectiveness among Canadian patients with PsA in their first year of treatment.
METHODS
CanSpA Research Network. This was a retrospective, registry-based, real-world study of patients with PsA in Canada within the CanSpA network, who were newly prescribed SEC as part of routine clinical care. The CanSpA network includes several registries in Canada of patients with SpA, each contributing data collected through routine clinical practice from multiple rheumatology clinics.
Data sources. This study used deidentified patient-level data of patients diagnosed with PsA extracted from 1 of 3 registries in the CanSpA network: the University Health Network (UHN) in Toronto, Ontario; Rhumadata in Montreal, Quebec; and Psoriatic Arthritis Comorbidities (PAC) in St. John’s, Newfoundland. Variables for data extraction were prespecified in a protocol and statistical analysis plan, and deidentified data were securely transferred to a common server where source data were then pooled and synthesized for analysis. All final analyses were conducted on the final pooled anonymized dataset.
Ethics approval. Ethics approval was received from the following review boards for this study: UHN Research Ethics Board (no. 20-5991, UHN), Advara (no. Pro00045670, Rhumadata), and Newfoundland and Labrador Health Research Ethics Board (no. 2020.302, PAC). A waiver of consent was granted for this study, as the study was a secondary analysis of deidentified patient data that posed no more than minimal risk to participants.
Patient population. To be eligible for inclusion, patients must have had a documented diagnosis of PsA and received treatment with SEC at any point between the date of SEC availability in Canada, including availability through clinical trials, and 6 months prior to the date of data extraction (ie, November 2021). Patients had to be 18 to 65 years of age at SEC initiation, with a minimum of 6 months of follow-up since the start of SEC treatment, regardless of treatment status. No additional exclusion criteria were applied.
Study variables. The primary outcome of the study was drug retention. Secondary outcomes included the following: proportion of patients in clinical remission or with low disease activity 12 months after SEC initiation (ie, baseline), average change in disease activity from baseline to 12 months, and reasons for discontinuation of SEC.
Retention was defined as the time from SEC initiation to SEC discontinuation. It was reported as the Kaplan-Meier (KM) estimate of the probability of being retained on SEC 12 months after initiation.
Remission and change in disease activity were assessed using several clinical indices: the Psoriatic Arthritis Disease Activity Score (PASDAS), the Disease Activity Index for Psoriatic Arthritis (DAPSA), the DAPSA in 28 joints (DAPSA28), the Disease Activity Score in 28 joints (DAS28) using C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), the Psoriasis Area and Severity Index (PASI), the physician global assessment, the patient global assessment, tender joint count using a 68- and 28-joint denominator (TJC-68 and TJC-28, respectively), and swollen joint count using a 66- and 28-joint denominator (SJC-66 and SJC-28, respectively). DAS28 and DAPSA28 indices were included for analysis because not all centers recorded 66- and 68-joint counts for disease activity assessment.
Remission was calculated as the proportion of patients falling below an index-specific threshold by 12 months, out of all patients with a nonmissing value for the respective index. The thresholds used were 1.9 for the PASDAS,18 4 for the DAPSA28,19 and 2.6 for the DAS28-CRP and DAS28-ESR.20 The PASDAS remission rate was reported both unadjusted and Lund Efficacy Index (LUNDEX)-adjusted, which accounts for both clinical response and the proportion of patients retained on SEC at 12 months.21 Change in disease activity, herein referred to as clinical response, was calculated as the mean percentage change in clinical index score from baseline to 12 months for patients with a score recorded at both timepoints.
In addition to the above outcomes, baseline demographics, clinical characteristics, and treatment history at baseline were also measured and reported.
Statistical analysis. The primary outcome, retention at 12 months, was reported as the KM estimate of the probability of being retained on SEC 12 months after initiation. Retention was reported overall and stratified by prior b/tsDMARD experience (experienced or naïve), sex (male or female), and duration of disease (0-3 years, 4-9 years, or ≥ 10 years). Comparison of retention between strata was performed using the log-rank test. Descriptive statistics for continuous variables are reported as mean and SD, and categorical variables are described as counts and proportions (%). All analyses were performed using SAS (version 9.4; SAS Institute).
RESULTS
Of the 276 patients identified in the CanSpA network who had a documented diagnosis of PsA and had been prescribed SEC, 213 were between 18 and 65 years of age at SEC initiation and had at least 6 months of follow-up. These 213 patients comprise the full analysis set (Figure 1). The mean age of the cohort at SEC initiation was 49.6 (SD 10.8) years, and 107 (50.2%) patients were male. The average BMI, calculated as weight in kilograms divided by height in meters squared, in the cohort was 30.8 (SD 6.5), and 57 (27.3%) patients were smokers. At baseline, 66 (31%) patients were b/tsDMARD-naïve vs 147 (69%) with prior b/tsDMARD experience, and the average time since PsA diagnosis was 9.1 (SD 8.9) years (Table 1).
Patient selection flowchart. PsA: psoriatic arthritis.
Baseline demographics and clinical characteristicsa.
Twelve-month retention. At 12 months postinitiation, 156 patients remained on SEC, with a KM-estimated retention rate of 73.6% (Figure 2). KM-estimated retention was 81.8% (54/66) among b/tsDMARD-naïve patients vs 69.9% (102/147) among b/tsDMARD-experienced patients, though this difference was not statistically significant (P = 0.19; Table 2). There was also no significant difference in retention rates between males (83/107, 79.3%) and females (73/106, 67.9%; P = 0.06). Retention was similar across disease durations, with a retention rate of 69.5% (54/79) for those with a disease duration of 0 to 3 years, 77.4% (42/53) for those with a disease duration of 4 to 9 years, and 75.1% (60/81) for those with a disease duration of 10 years or longer (P = 0.98).
Kaplan-Meier–estimated retention to SEC from SEC initiation up to 12 months postinitiation. PsA: psoriatic arthritis; SEC: secukinumab; Surv.: survival.
KM-estimated retention rate at 12 monthsa, stratified by b/tsDMARD experience, sex, and disease duration.
Clinical remission and response. Denominators for calculating clinical remission include only patients with a nonmissing value for the respective disease activity index at 12 months. In total, 10 out of 70 (14.3%) patients with a PASDAS score had achieved PASDAS-based remission by 12 months. The LUNDEX-adjusted PASDAS remission rate was 10.5%. A total of 17 out of 110 patients (15.5%) with DAPSA28 scores had achieved DAPSA28-based remission. In addition, 79 out of 124 patients (63.7%) and 67 out of 104 patients (64.4%) with DAS28-CRP and DAS28-ESR scores had achieved DAS28-CRP– and DAS28-ESR–based remission by 12 months, respectively.
Clinical response was calculated as the mean percentage change in clinical index score from baseline to 12 months. Denominators for calculating clinical response included only patients with nonmissing values for the respective disease activity index at both baseline and 12 months. Patients showed an overall trend of improvement in clinical scores from baseline to 12 months. Among patients with a score recorded both at baseline and 12 months, the total DAPSA (ie, based on 66- and 68-joint counts) improved by 36.5% (from 21.7 to 13.7) from baseline to 12 months. The DAS28-CRP and DAS28-ESR showed similar improvements, dropping by 20.9% (from 2.8 to 2.2) and 20.2% (from 2.9 to 2.3), respectively. With respect to PsO, the PASI improved by 65.8%, dropping from 6.4 at baseline to 2.2 at 12 months. The largest improvements were seen for TJC-68 and SJC-66, improving by 65.5% (from 6.0 to 2.1) and 73.7% (from 3.8 to 1.0), respectively, from baseline to 12 months. Full results for all recorded clinical indices are described in Figure 3.
Mean clinical response from baseline to 12 months for patients with both a baseline and a 12-month score recorded. The 12-month clinical outcome scores were recorded as the value recorded closest to the date 12 months after SEC initiation and had to occur within 4 months of the 12-month date. In order to calculate the change from baseline to 12 months and be considered nonmissing, patients had to have a score recorded at both baseline and 12 months. CRP: C-reactive protein; DAPSA: Disease Activity Index for Psoriatic Arthritis; DAS28: Disease Activity Score in 28 joints; ESR: erythrocyte sedimentation rate; PASDAS: Psoriatic Arthritis Disease Activity Score; PASI: Psoriasis Area and Severity Index; PGA: physician global assessment; PtGA: patient global assessment; SEC: secukinumab; SJC-28: swollen joint count-28 joints; SJC-66: swollen joint count-66 joints; TJC-28: tender joint count-28 joints; TJC-68: tender joint count-68 joints.
Reasons for discontinuation. Of the 213 patients aged 18 to 65 years that had initiated SEC at baseline and had a minimum of 6 months of follow-up, 52 (24.4%) had discontinued SEC by 12 months (ie, had a documented discontinuation date before 12 months or before being lost to follow-up). Of these 52 patients, 35 (67.3%) cited lack or loss of effectiveness as their reason for discontinuation, with 22 discontinuing at or before 6 months and 13 discontinuing after 6 months of SEC treatment. Other reasons for discontinuation included adverse events (8/52, 15.4%), patient preference (1/52, 1.9%), or another reason not further specified (5/52, 9.6%), with no reason provided from the remaining 3 (5.8%) patients (Table 3).
Reasons for discontinuation among those discontinuing SEC by 12 months.
DISCUSSION
This registry-based real-world study of SEC for the treatment of PsA—1 of the first of its kind in Canada and the first endeavor for the CanSpA Research Network—found that 73.2% of patients with PsA remained on SEC 12 months after initiation and observed an improvement in clinical PsA symptoms over the same period. These results, which included both b/tsDMARD-naïve and -experienced patients, provide additional evidence that SEC is effective and well tolerated for the treatment of PsA in a Canadian real-world setting, and further substantiate findings from other real-world studies on SEC internationally. Results published by the EuroSpA Research Collaboration Network in Europe, for example, found a similar retention rate of 76% 12 months after initiating SEC treatment,16 and a US claims database study estimated 12-month persistence to SEC to be roughly 64%.22 Despite slightly lower retention observed in the US real-world study, SEC still showed the lowest discontinuation rate and highest persistence and adherence relative to all other biologics analyzed in the claims database. The rates of retention observed in this study and others16,17,23 are an indication that continued treatment with SEC is beneficial for the majority of patients with PsA in a real-world setting.
In addition to the observed retention, we also observed improvements in clinical scores after 12 months of treatment across all clinical indices assessed in this study. DAPSA scores showed average improvements of 37.9%, while TJC-68 and SJC-66 improved by more than 65%. As study populations vary in their baseline characteristics and disease activity, and choice of methodology and clinical indices to measure effectiveness can be inconsistent across different studies, comparisons to other real-world studies and clinical trials are not necessarily straightforward and should be made cautiously. In the FUTURE 5 trial, the DAS28-CRP improved by 1.29 and 1.49 points after 16 weeks of treatment for those on a 150-mg dose and 300-mg dose, respectively. This trend persisted to 52 weeks, with patients on 150-mg doses showing reductions of 1.72 to 1.83 points from baseline, and those on a 300-mg dose showing a reduction of 1.92 from baseline. However, there are key differences in methodology that must be considered when comparing efficacy observed in clinical trials vs effectiveness observed in a real-world setting. To start, patients in the FUTURE 5 trial assigned to the placebo group also experienced an improvement in the DAS28-CRP of 0.63 points at 16 weeks, suggesting that the sole therapeutic effect of SEC may actually be closer to 0.66 to 0.86 at 16 weeks, and this value was more comparable to the 0.60-point improvement seen here. Patients receiving SEC in a clinical trial setting are also not subject to the same criteria for reimbursement of SEC as in a real-world setting. For example, Canadian patients must typically demonstrate failure on at least 2 DMARDs to be eligible for a biologic under a public drug plan. Given these criteria and the relatively recent availability of SEC, patients receiving SEC in a real-world setting are more likely to have previously failed conventional and other bDMARDs, relative to patients enrolled in clinical trials, and may, therefore, be less likely at baseline to show adequate response to treatment. In addition, stricter eligibility criteria employed in the FUTURE 5 trial, indeed in many clinical trials, may have resulted in a sample of patients with more severe disease at baseline who are more likely to show improvements with treatment. Patients in the FUTURE 5 trial had baseline DAS28-CRP scores over 4.5 relative to 2.8 in CanSpA, and they had TJC-68 of roughly 20 compared to 6 reported here. Importantly, the DAS28-CRP itself may also be limited in its ability to measure improvement in disease activity for PsA, as it does not account for manifestations in the lower limbs, and an index such as the DAPSA or TJC-68/SJC-66 may be more sensitive in detecting improvement. Nevertheless, patterns of DAPSA28-based remission observed in CanSpA appeared highly similar to, or even an improvement on, other real-world studies of SEC. Around 16% of patients in CanSpA had achieved DAPSA28-based remission by 12 months, relative to 11% in EuroSpA. In addition, DAS28-CRP–based remission rates were 64% in CanSpA vs 39% in EuroSpA,16 though the use of 28-joint count indices may overestimate remission, as they omit a large number of joints relevant for measuring disease activity. However, baseline differences in populations must still be considered. Although both CanSpA and EuroSpA used a real-world population with minimal eligibility criteria in a clinical setting, some differences did exist between the groups at baseline. Specifically, the EuroSpA cohort had a smaller proportion of b/tsDMARD-naïve and male patients, which could have contributed to baseline differences in the likelihood of treatment success.
Although we see some inevitable variability in estimates of clinical effectiveness across different populations and methodologies, overall trends still appear compatible across studies. That is, clinical signs and symptoms of PsA show consistent improvements with SEC treatment over at least 12 months, with a considerable proportion of patients achieving disease remission status across several clinical indices, particularly those that account for lower limb manifestations, such as the DAPSA, TJC-68, and SJC-66. Although the DAS28 also shows improvement over at least 12 months, we reassert that it is not as sensitive as other measures, such as the DAPSA, for the measurement of PsA disease activity because it misses lower extremity joints that are commonly affected in PsA, as evidenced by the greater improvements seen for TJC-68/SJC-66 vs TJC-28/SJC-28. Indeed, the TJC-68/SJC-66 are the first fully endorsed outcome measurements by GRAPPA-OMERACT for their sensitivity in identifying the heterogeneous disease manifestations of PsA.24
Overall, the improvements in clinical scores and rates of remission observed in this analysis, in addition to the high rates of retention observed at 12 months, are indicative of treatment success with SEC among Canadian patients with PsA. These results are corroborated by findings from clinical trials and additional studies in real-world settings. However, some differences between the current results and other literature and clinical expectations should be noted. Specifically, our study found no statistically significant difference in retention between b/tsDMARD-naïve and b/tsDMARD-experienced patients. Clinically, one would expect patients with previous failure on biologic treatment to be at increased risk of additional treatment failure and subsequent discontinuation, and this pattern has been previously shown in a number of real-world studies, including those of SEC for PsA and other SpA diseases, as well as other biologic classes for PsA and SpA.16,17,22,23 Direct comparisons of b/tsDMARD-experienced vs -naïve patients have yielded higher retention to IL-17 inhibitors among naïve patients.16,17,22 In addition, cohorts comprising predominantly patients refractory to biologic therapy, such as that of a real-world study in Spain, showed lower retention rates relative to other cohorts.23 Similarly, other real-world studies have found male sex to be protective against discontinuation of SEC,25 whereas our analysis found no statistically significant difference in retention between males and females. That said, our results did trend toward higher retention among b/tsDMARD-naïve vs -experienced patients (81.8% vs 69.9%) and among males vs females (79.3% vs 67.9%); the lack of statistical difference between these groups may be mostly reflective of sample sizes.
This study was one of the first real-world studies in Canada to describe real-world retention of SEC in patients with PsA, enhancing the evidence base for PsA and providing insight into how treatment performs outside of clinical trial settings. However, use of real-world data does confer some limitations. First, due to limited capture, this study used modified disease activity indices such as the DAPSA28 to assess clinical remission and response for PsA, which has not been validated using a 28-joint count. Similarly, data required to adequately evaluate changes in PsO presentation and severity were not available. For example, body surface area was not captured for any patients, and capture for the PASI was quite low. This was a major limitation in understanding the real-world effectiveness of SEC for skin manifestations, for which it has previously demonstrated excellent efficacy in clinical trials.12 Beyond the low capture for the PASI, there was also substantial data missingness across other variables and outcomes, particularly for other clinical indices, which could serve to bias the results if data are more likely to be missing for patients showing a certain response (eg, for patients with low disease activity). Though challenges with data collection and missingness are common among real-world studies relative to clinical trials, they are nevertheless limitations, and more thorough, routine data collection may be a valuable target for future real-world evidence generation endeavors. Further, limited sample sizes, particularly for b/tsDMARD-naïve patients, may have limited the ability to conduct meaningful subgroup analyses and may have resulted in underpowered statistical comparisons. Finally, this study used retrospective data primarily collected for clinical practice from the contributing databases, which may not have rigorous protocols or validation practices. As such, variation in both clinical management and data collection practices across rheumatologists in the CanSpA network could introduce inconsistencies in the data and uncertainty about results.
Despite all of these limitations, this study has begun to fill an important knowledge gap in Canada for the real-world use and effectiveness of SEC for PsA. This study supports the effectiveness of SEC in the treatment of PsA in a Canadian real-world setting, but also underscores the need for additional studies investigating predictors of SEC response and comparisons with other treatments. Beyond this treatment-focused study, the CanSpA Research Network represents an opportunity to continue building and improving the real-world evidence base for treatment of SpA in Canada, and plans to broaden this initiative are already underway. Future studies within the CanSpA network will aim to improve upon the limitations discussed above and continue addressing real-world rheumatology evidence needs in Canada.
Footnotes
This study was sponsored by Novartis Pharmaceuticals Canada Inc.
DDG has received research grant support from AbbVie, Amgen, Celgene, Eli Lilly, Janssen, Novartis, Pfizer, and UCB, and consulting fees from AbbVie, Amgen, Celgene, Eli Lilly, Galapagos, Gilead, Janssen, Novartis, Pfizer, and UCB. DC has received research grant support from AbbVie, Amgen, Eli Lilly, Fresenius-Kabi, Novartis, Pfizer, Sandoz, Teva Pharmaceuticals, and Sanofi-Genzyme, and consulting or speaker fees from AbbVie, Amgen, Eli Lilly, Novartis, Pfizer, Sandoz, Celltrion, and Teva Pharmaceuticals. MK has received consulting fees from AbbVie, Amgen, Gilead, Novartis, Pfizer, and UCB. RDI has received consulting fees from AbbVie, Janssen, Lilly, Novartis, and Sandoz. PL and SH are employees of Novartis Pharmaceuticals Canada Inc. DN is an employee of IQVIA. IQVIA received consulting fees from the sponsor, Novartis Pharmaceuticals Canada Inc., for the study.
- Accepted for publication November 23, 2022.
- Copyright © 2023 by the Journal of Rheumatology
